GLACIATION OF PRINCE WILLIAM SOUND 477 served up to an elevation of 200 feet and glacial lakes (PI. CLXXXIE) nearly up to 400 feet, above which the absence of erratics showed that the area has not been ice-covered. The highest glaciated rock ledges seen were at an elevation of 185 feet, and these were strongly polished and striated. Montague Island may also be partly driftless, though not for so large an area as Hinchinbrook Island, which had fewer local glaciers. The rolling topography of the plateau-like southwestern end of Hinchinbrook Island continues throughout most of the southern half of the island. From sea level in Prince William Sound and on the Pacific Ocean it closely simulates a glacially-smoothed topography (PL CLXXXIIL); but it is absolutely certain that it has never been covered by ice. It seems to be a part of the warped, southward-tilted peneplain, described by Spencer* and jay Grant and Higgins,8 and to owe its rounded topography to sub-aerial denudation. In it a few cirques of small local glaciers are cut. The Glaciation of the Continental Shelf. In the Gulf of Alaska opposite the mouth of Prince William Sound there is an exceptionally wide continental shelf. Mention has already been made of the fact it is necessary to go 40 to 70 miles from shore before a depth of 600 feet is reached and that in the 56 miles between Hinchinbrook Entrance and Middleton Island the depth of water is only 200 to 450 feet.8 The question naturally arises as to whether glaciation has played any part in producing this broad, shallow continental shelf. The shelf, itself, is, of course, a feature which existed before and is independent of glaciation. That it may have been glacially modified, however, is suggested by (1) observations at Hinchinbrook lighthouse and at the southern end of Elrington Island which prove that the glacier from Prince William Sound once extended out into the the Pacific Ocean; (£) the deposits upon Middleton Island, which Dawson4 has interpreted as glacial till. This suggests that the glacier moved outward across the continental shelf to or beyond Middleton Island, and that the Gulf of Alaska once had a great, tidal ice front similar to the barriers of the Antarctic region, but not floating because the water was too shallow. It is difficult to determine whether the expanded glaciers actually pushed out to sea beyond the islands, glaciating the continental shelf. Aside from the glaciated condition of Hinchinbrook Entrance, there are three facts suggesting that they did. In the first place there is no visible deposit which can be considered representative of the terminal accumulation of this long period of expanded piedmont glacier in Prince William Sound. The inference seems warranted that it lies beneath the sea on the continental shelf. A second fact, suggesting the correctness of this conclusion, is the shallowing of the water south of the entrance to the sound. Soundings are not numerous enough to warrant definite conclusions on this point, but beyond the deep entrance to the sound between Montague and Hinchinbrook Islands, there is a broad area of shallower water with soundings of from 120 to 240 feet and beyond that only slightly deeper water. More detailed soundings are necessary before the significance of this can be definitely determined. A third noteworthy point is the depth of the entrance between the two islands. Its depth i Spencer, A. C., Bull. Geol. Soc. Amer., Vol. XIV, 1903, pp. 117-182. »Grant, U. S. and Biggins, D. F., Bull. 448, U. S. Geol. Survey, 1910, p. 15. • See Chart 8502, U. S. Coast and Geod. Survey. « Dawson, G. M., Notes on the Geology of Middleton Island, Alaska, Bull. Geol. Soc. Amer., Vol. 4,1893, pp. 427-431; Brooks, A. H., Bull. 542, U. S. Geol. Survey, 1913, p. 48.